Comparison of PCR-based DNA fingerprinting techniques for the

Document Sample
Comparison of PCR-based DNA fingerprinting techniques for the Powered By Docstoc
					 International Journal of Systematic Bacteriology (1 998),48,127-1 39                                                                                                                                Printed in Great Britain

                                                                                         Comparison of PCR-based DNA fingerprinting
                                                                                         techniques for the identification of Listeria
                                                                                         species and their use for atypical Listeria
                                                                                         Mario Vaneechoutte,’ Patrick Boerlin,’t Hans-VoIker Tichyt3
                                                                                         Elizabeth Bannerman,’ Birgit Jager3 and Jacques Bille’

                                                                                         Author for correspondence: Mario Vaneechoutte. Tel: +32 9 240 36 92. Fax:                              + 32 9 240 36 59.
                                                                                         e-mail : Mario .Vaneechoutte @rug.

1     Department of                                                                     Four PCR-based DNA fingerprinting techniques were compared for their ability
      Microbiology and                                                                  to identify a t the species level a heterogeneous collection of isolates
      Immunology, Blok A,
      University Hospital, 9000                                                         belonging to the six valid Listeria species. 16s rDNA-RFLP analysis identified all
      Gent, Belgium                                                                     species and 16s rDNA-SSCP analysis identified almost all species. Also, isolates
* Swiss National Center for                                                             with unusual biochemical characteristics and/or unusual antigenic composition
      Listeriosis and WHO                                                               could be identified correctly. rRNA-intracistronic length polymorphism analysis
      Collaborative Center for                                                          suffered from high intraspecific variability, a limited number of fragments per
      Foodborne Listeriosis, 1011
      Lausanne, Switzerland                                                             profile, and small length differences between the spacers of different species.
                                                                                        tRNA-intergenic length polymorphism analysis resulted in identification of all
3    TUV Energie und
     Systemtechnik GmbH,                                                                isolates but one, when fluorescent DNA capillary electrophoresis was used
     Abteilung Biologische                                                              such that fragment length differences of 1 bp could be resolved. The four
     Sicherheit, Freiburg i. Br.,                                                       techniques yielded comparable results relevant to the taxonomy of Listeria.
                                                                                        They all indicate a high degree of genetic relatedness between L. innocua and
                                                                                        L. welshimeri, homogeneity of L. grayi, distinct but clear relatedness of L. grayi
                                                                                        to the other Listeria species, a clear distinction between the two subspecies of
                                                                                        L. ivanovii, and a clear distinction between Listeria isolates and isolates from
                                                                                        closely related taxa or from species which are phenotypically difficult to
                                                                                        distinguish from Listeria. New sequence determination of t h e 16s rRNA gene
                                                                                        was necessary to obtain sequences in accordance with the findings of 165
                                                                                        rDNA-RFLP analysis.

                                                                                                Keywords: Listeria, PCR-based DNA fingerprinting, 16s rDNA-RFLP, 16s rDNA-
                                                                                                          SSCP, spacer length polymorphism

INTRODUCTION                                                                                                                                                   ivanovii, L. innocua, L . welshimeri, L. seeligeri and L.
                                                                                                                                                               grayi). The former species ‘ L . murrayi’ has been
The genus Listeria is very homogeneous, except for L .                                                                                                         assigned to the species L. grayi (33). Only L. mono-
grayi which is clearly distinct from the other species (4,                                                                                                     cytogenes and L. ivanovii are recognized as significant
7,22, 30, 34). The taxonomy of the genus Listeria has                                                                                                          pathogens of humans and animals. The Listeria species
been difficult to assess in the past, but molecular                                                                                                            differ in only a few phenotypic characteristics (12, 22,
methods (4, 30, 32) have clearly demonstrated the                                                                                                              31) and this high degree of similarity is the source of a
existence of six Listeria species (L. monocytogenes, L.                                                                                                        lot of confusion in their identification. Furthermore,
                                                                                                                                                               several other species, such as Bacillus species, Entero-
......................I..............   1   ...................................................................................    .
                                                                                                                                  I.   .....................   coccus species (particularly Enterococcus faecalis)
t Present address: Department of Pathobiology, University of Guelph,
                                                                                                                                                               and Cellulomonas (Oerskovia) turbata, are often
Guelph, Ontario, Canada NIG 2W1.                                                                                                                               misidentified as Listeria species by inexperienced
Abbreviations: ARDRA, amplified rDNA restriction analysis; SSCP, single                                                                                        microbiologists. Finally, the type strain of L. mono-
strand conformational polymorphism.                                                                                                                            cytogenes is phenotypically abnormal (23) and would
The EMBL accession numbers for the sequences reported in this paper are                                                                                        be difficult to identify if it was encountered as a field
X9852CX98532.                                                                                                                                                  isolate.

00576 0 1998 IUMS                                                                                                                                                                                                       127
M. Vaneechoutte and others

Methods which combine the reliability of genotypic                  supernatant of this boiled cell suspension was used as the
identification with the speed and simplicity of pheno-              target DNA and was added - using filter-protected tips - to
typic methods such as biochemistry and serology,                    45 pl aliquots of PCR mix for all DNA fingerprinting
would be more than welcome, particularly in the                     techniques.
presence of biochemically and serologically atypical                Amplification mixture. All reaction mixtures contained - at

isolates. Four PCR-based DNA fingerprinting                         a final reaction volume of 50 pl - 0.005 U Goldstar poly-
methods have been described to enable identification                merase pl-l (Eurogentec), 100 pM each dNTP (Pharmacia
of cultured bacterial organisms to the species level on             Biotech) and 0.2-0.5 pM each primer in reaction buffer
the basis of genotype: rDNA-RFLP analysis (or                       [7.5 mM Tris/HCl, 2.5 mM MgCl,, 20 mM (NH,),SO,, 0.1 %
A R D R A , amplified rDNA restriction analysis) and                (w/v) Tween 20, pH 9-01.After aliquotting in 45 pl volumes,
                                                                    mixtures were overlaid with 40 pl mineral oil.
rDNA-single-strand conformational polymorphism
(SSCP) analysis, which both make use of the species-                ARDRA. The 16s rRNA gene (16s rDNA) was amplified
specific sequence information in the 16s rRNA gene;                 using 0.2 pM primers complementary to conserved regions
and rDNA-spacer-PCR and tDNA-spacer-PCR,                            present at the edges of the 16s rDNA. The sequences of the
which have been reported to provide species-specific                primers were S'TGGCTCAGATTGAACGCTGGCGGC
                                                                    (Escherichia coli nucleotide positions 10-27) and 5'TAC-
length polymorphism of rRNA intracistronic and                      CTTGTTACGACTTCACCCCA (E. coli nucleotide pos-
tRNA-intergenic spacer regions, respectively. In this               itions 1507-1485). After initial denaturation at 95 "C for
study we assessed the suitability of these four methods             5 min, the reaction mixtures were cycled 35 times for 45 s at
for the identification of the species of the genus                  95 "C, 45 s at 55 "C, and 1 rnin at 72 "C. Finally, a 7 rnin
Listeria.                                                           extension period at 72 "C was carried out. The presence and
                                                                    yield of specific PCR products were determined by Multi-
ARDRA has been studied most thoroughly and has                      Purpose (Boehringer) agarose (1 %, w/v) - ethidium bro-
been used for the identification of species of different            mide (50 ng ml-l) gel electrophoresis for 30 rnin at 7 V cm-l.
taxa, e.g. Acinetobacter (39), Clostridium (37), Coma-              Ten microlitre aliquots of the amplified DNA were used
monadaceae (4 l), Corynebacterium (40), Mollicutes                  without further purification for restriction digestion. The
(9), Mycobacterium (38), Leptospira (29) and Strep-                 following enzymes were used: AluI, AvaI, BanI, BfaI,
tococcus (19), and to study phylogenetic relationships              Bsp12861, BstEII, BstUI, CfoI (HhaI isoschizomer), DraI,
in Bacillus (16), Clostridium (13), Moraxella (18) and              EcoT381, HaeIII, HinfI, MboI, MspI, MvaI, NciI, RsaI,
Xanthomonas (28).                                                   Sau961, ScrFI, StyI, TaqI and Tru9I. Restriction digestion
                                                                    was carried out for 1 h at 37 "C, except for TaqI and Tru91,
Widjojoatmodjo et al. (44, 45) used SSCP analysis of                which were incubated at 65 "C, in 20 pl volumes of com-
part of the 1 s rDNA to identify a number of species,               mercial incubation buffer containing 1 U p1-l restriction
including all six Listeria species. Jensen et al. (20)              enzyme and 5 pl amplified 16s rRNA gene product. Re-
showed that rDNA-spacer-PCR could be used to                        striction fragments were separated by gel electrophoresis for
identify Escherichia species, some Enterobacter and                 3 h at 7 V cm-l on 3 % Metaphor agarose (FMC Bio-
Staphylococcus species and five of the six Listeria                 Products Europe) containing 50 ng ethidium bromide ml-'.
species (L. seeligeri was not studied). Others used this            165 rDNA-SSCP analysis. rDNA-SSCP analysis was carried
approach successfully for the identification of Legion-             out essentially as described by Widjojoatmodjo et al. (44).
ellaceae (17).                                                      Primers P l l P (5'GAG GAA GGT GGG GAT GAC GT)
                                                                    and P13P (5'AGG CCC GGG AAC GTA TTC AC) were
tDNA-spacer-PCR has been used for the identification                used at concentrations of 0.2 pM to amplify a 216 bp
of Staphylococcus species (42, 43), Streptococcus                   fragment of the V6 region of the 16s rDNA ( E . coli 16s
species (26) and Acinetobacter species (10, 46).                    rDNA positions 1 175-1 390). After initial denaturation at
                                                                    95 "C for 5 min, the reaction mixtures were cycled 35 times
METHODS                                                             for 20 s at 95 "C, 1 rnin at 55 "C and 1 rnin at 72 "C. Finally,
                                                                    a 7 rnin extension period at 72 "C was carried out: 0-3 pl
Bacteria. Table 1 lists the isolates which were used. Besides       volumes of the amplification products were mixed with 6 pl
representatives of all the six valid Listeria species, including    water and 3 pl sequencing buffer (95 YO    formamide, 20 mM
the two subspecies of L. ivanovii (3), isolates of the species L.   EDTA, 0.05% bromophenol blue and 0.05% xylene
monocytogenes, L. innocua, L. welshimeri and L. ivanovii            cyanol). The mixture was heated to 95 "C for 5 min and
representing unusual biotypes or atypical antigenic com-            loaded on the gels. The amplification products were separ-
positions were also included and tested blindly. The Listeria       ated in non-denaturing 0.5 x MDE gels (FMC BioProducts)
isolates representing atypical biochemical or serological           containing 5 % glycerol. The gels were run overnight in a
reactions were also tested with API-Listeria galleries              Protean I1 xi electrophoresis cell (Bio-Rad) in 0.5 x TBE at
(BioMerieux), following the recommendations of the manu-            room temperature at 90 V. The gels were stained using the
facturer. Isolates of genera phylogenetically closely related       Silver Stain Plus kit (Bio-Rad) and the profiles were
to Listeria were also included since these may be pheno-            interpreted visually.
typically misidentified as Listeria.                                rRNA-intracistronic spacer length polymorphism analysis
Sample preparation for PCR. Target DNA was prepared from            (rDNA-spacer-PCR). The intracistronic rDNA spacer regions
cells grown overnight at 37 "C on Mueller-Hinton Agar I1            were amplified by using primers - at concentrations of
(MHAII ; BBL Microbiology Systems) supplemented with                0.5 pM - flanking the 16s-23s rDNA spacer region (for-
5 % sheep blood. A 1 p1 volume of cells was suspended in            ward, 5'GAA GTC GTA ACA AGG; reverse, 5'CAA GGC
300 pl 10 mM Tricine/l.5 mM MgC1, and heated at 100 "C              ATC CAC CGT) (20). After initial denaturation at 95 "C for
for 10 rnin in a heating block. Five microlitres of the             5 min, the reaction mixtures were cycled 25 times for 1 min

128                                                                              International Journal of Systematic Bacteriology 48
                                                                  PCR fingerprinting for identification of Listeria species

at 94 "C, 2 rnin at 55 "C, and 2 rnin at 72 "C, followed by a
final 7 rnin extension period at 72 "C. Twenty microlitres of
each amplification product was separated in 4 % poly-
acrylamide gels using 1 x TBE at constant voltage in a
Protean I1 xi electrophoresis cell. The gels were stained in a
1 pg ethidium bromide ml-I solution and photographed
under short wavelength UV light. The migration profiles
were compared visually.
tRNA-intergenicspacer length polymorphismanalysis (tDNA-
spacer-PCR). tDNA-spacer-PCR was carried out essentially
as described by Welsh & McClelland (42). Primers T5A
(5'AGG CCG CGG GTT CGA ATC C) were used at
concentrations of 0.5 pM to amplify the spacer regions
between the tRNA genes. For analysis by capillary electro-
phoresis the T3B primer was labelled at the 5'end with the
fluorescent amidite TET (Applied Biosystems). After initial
heating at 95 "C for 7 min, reaction mixtures were cycled 40
times for 30 s at 95 "C, 30 s at 50 "C and 2 rnin at 72 "C,
followed by final extension for 7 rnin at 72 "C. Capillary
electrophoresis of fluorescently labelled PCR products was
done on an ABI 310 (Applied Biosystems) using the POP-4
polymer and capillaries and with the TAMRA-500 standard
as the internal lane size standard.
165 rDNA sequence determination. An approximately
 1.45 kb segment of the 16s rRNA gene was amplified using
primer 41f (5'GCT CAG ATT GAA CGC TGG CG) and
the biotin-labelled primer 14881-(S'biotin-CGG TTA CCT           Fig, 1. 165 rDNA-RFLP profiles obtained by 3 % agarose gel
                                                                 electrophoresis of digests of the amplified 165 rDNA after
TGT TAC GAC TTC ACC). Forty microlitres of the PCR               restriction with Alul .
sample were added to Dynabeads M280 Streptavidin
(Dynal), the DNA was denatured and the biotin-labelled
single strand was immobilized onto the beads and used as
one template for the sequencing reaction using fluorescent       BstUI, CfoI, RsaI and Tru9I enabled some differen-
primers 358f (5' AGA CTC CTA CGG GAG GCA GCA                     tiation (Table 1). Combination of the profiles observed
GT), 536f [5' GTG CCA GC(AC) GCC GCG GTA ATA C]                  after digestion with AluI (Fig. 1) and BstUI enabled
and 928f [5' TAA AAC T(CT)A AA(GT) GAA TTG ACG                   identification of all Listeria species, except for L.
GGG]. The opposite strand remaining in solution was also         innocua and L. welshimeri, for which a constant
used as template using fluorescent primers 336r [5' ACT          difference could be shown only after restriction of the
GCT GCS (CT)CC CGT AGG AGT CT] and 51.51-[5'                     16s rDNA with Tru9I (Table 1, Table 2). Also, only
G(AT)A TTA CCG CGG C(GT)GCTG GCA C].                             minor but constant differences could be observed for
Sequencing reaction products obtained by using the T7            BstUI restriction profiles of L. ivanovii subsp. ivanovii
Autoread kit (Pharmacia Biotech) were processed using an
ALF DNA sequencer (Pharmacia Biotech) and the ALF                and L . ivanovii subsp. londoniensis, and for L.
manager software. DNA sequence comparisons were done             welshimeri and L . seeligeri (Table 2). L. grayi was
using the HIBIO DNASIS software package (Hitachi).               clearly different from the other five species, since
Published Listeria sequences were obtained from the EMBL         restriction with AluI and RsaI yielded unique re-
database and the database included in the ARB package (see       striction profiles not observed for any other Listeria
below). Tree construction was done using the ARB software        species (Table 1, Table 2). One L. innocua isolate
package developed by Oliver Strunk, Wolfgang Ludwig and          (TI6520) had a BstUI profile (lc) with an additional
others from the Microbiology Department of the Technical         250 bp fragment. For L. monocytogenes isolate I P l l
University, Munich, Germany. The software version for            an extra fragment was observed repeatedly in the CfoI
Linux and the database (6aug9.ascii.arb) were obtained by        profile. This could be resolved as the result of in-
anonymous ftp from
                                                                 terference of minor non-specific amplification, not
Computer-assisted restriction analysis of the 16s rRNA           present in one of the collaborating laboratories. The
gene was carried out with HIBIO DNASIS on EMBL                   isolates of the closely related non-Listeria species could
sequences X56148-X56154 (7) and on the sequences ob-             easily be identified as such by restriction analysis of the
tained in the present work. Distance values were calculated
using the neighbour-joining method with Kimura correc-           16s rDNA, although some of the restriction enzymes
tion.                                                            yielded identical profiles for some Listeria and non-
                                                                 Listeria species (Table 1).
                                                                 16s rRNA gene sequence determination and
ARDRA                                                            phylogenetic tree construction for the genus Listeria
Restriction of the 1 s rRNA gene was carried o u t with          T h e sequences of the 16s rRNA genes were determined
a total of 22 enzymes. Only restriction with AluI,               for seven isolates (indicated in Table 1). The distance

International journal of Systematic Bacteriology 48                                                                     129
M. Vaneechoutte and others

Table 1. Designation of isolates, serotype and biotype, and of profiles observed for different PCR-based DNA
fingerprinting techniques for species identification

I   Code
                  Original no.                  Species*                    Serotype or unusual
                                                                          agglutination reactionst
                                                                                                     AluI      Cfol

                                                                                                                           RsaI       BstUI   Tru9I
                                                                                                                                                                     PCR    1

     1     LL1955                  L. monocytogenes                  4b                               1            1        1          la      la         1             6
     2     ZH7415                  L. monocytogenes                  1/2a                             1            1        1          la      la         1             4
     3     SLCC2371                L. monocytogenes                  1/2a                             1            1        1          la      ND         1             4
     4     SLCC2755                L. monocytogenes                  1/2b                             1            1        1          la      ND         1             5
     5     SLCC2372                L. monocytogenes                  1/2c                             1            1        1          la     ND          1             4
     6     SLCC2373                L. monocytogenes                  3a                               1            1        1          la     ND          1             4
     7     SLCC2540                L. monocytogenes                  3b                               1            1        1          la      ND         1             4
     8     SLCC2479                L. monocytogenes                  3c                               1            1        1          la      ND         1             4
     9     SLCC2374                L. monocytogenes                  4a                               1            1        1          la      la         1             9
    10     SLCC2375                L. monocytogenes                  4b                               1            1        1          la      ND         1             7
    11     SLCC2376                L. monoeytogenes                  4c                               1            1        1          la      ND         1            10
    12     SLCC2377                L. monocytogenes                  4d                               1            1        1          la      ND         1             4
    13     SLCC2378                L. monocytogenes                  4e                               1            1        1          la      ND         1             6
    14     SLCC2482                L. monocytogenes                  7                                1            1        1          la      ND         1             5
    58     IP11                    L. monocytogenes                  V/Vl ;VI; V1I;VIII ;A ;B; C      1            1       ND          la      la         1             8
    59     IP15                    L. monocytogenes"                 4b                               1            1       ND          la      ND         1            ND
    65     Nv3953                  L. monocytogenesb                 1/2b                             1            1       ND          la      ND         1            ND
    66     NV4 172                 L. monocytogenesb                 1/2a                             1            1       ND          la      ND         1            ND
    68     LU4526                  L. monocytogenesC                 4b                               1            1       ND          la      ND         1            ND
    77     BS6179                  L. monocytogenes"                 4b                               1            1       ND          la      ND         1            ND
    79     GE6366                  L. monocytogenes'                 1/2c                             1            1       ND          la      ND         1            ND
    15     SLCC3379/ATCC 330905    L. innocua                        6a                               2            1        1          la      la         2            13
    16     LL271                   L. innocua                        6b                               2            1        1          la      la         2            11
    17     VD8304                  L. innocua                        6a                               2            1        1          la      la         2            12
    18     8716                    L. innocua                        6b                               2            1        1          la      la         2            11
    25     VD7945                  L. innocua                        6b                               2            1        1          la      la         2            11
    69     GE5432                  L. innocua                        I; I/II;A; B;C (1/2b)            2            1       ND          la      la         2            ND
    70     GE5433                  L. innocua                        I; I/lI;A; B;C (1/2b)            2            1        1          la      la        ND            ND
    71     TI5796                  L. innocua                        V/VI ;VII ;XV; B; C ; D          2            1       ND          la      la         2            ND
    76     VD6 140                 L. innocua                        V/VI;VII;XV;B;C;D                2            1        1          la      la        ND            ND
    81     TI6520                  L. innocua                        XV ;B ;C ;D                      2            1       ND          lc      la         2            ND
    82     TI6551                  L. innocua                        V/VI ;VII ;XV: B ;C ; D          2            1        1          la      la        ND            ND
    83     BL6621                  L. innocua                        No somatic factors A;B;C         2            1        1          la      la        ND            ND
    87     BL6830                  L. innocua                        VII; VI1I;A; B; C                2            1       ND          la      la        2             ND
    23     SLCC5334                L. welshimeri                     6a                               2            1        1          la      lb        2             14
    24     ZH7199                  L. welshimeri                     6b                               2            1        1          la      lb       2+3            14
    60     LL244                   L. welshimeri                     V/VI ;VI ;VII ;X ;B ;C ;D        2            1       ND          la      lb       2+3            ND
    61     LL519                   L. welshimeri                     I;I/II; A; B; C (1/2b)           2            1       ND          la      lb        2             ND
    62     LL520                   L. welshimeri                     I;I/II;A;B;C (1/2b)              2            1       ND          la      lb        2             ND
    63     BS2094                  L. welshimeri                     V/VI;VI;VII;X;B;C;D              2            1        ND         la      lb       2+3            ND
    67     SO4352                  L. welshimeri                     V/VI;VI: VI1;X: B;C; D           2            1        1          la      lb        ND            ND
    72     TI5800                  L. welshimeri                     V/VI;VI;VII;X;B;C;D              2            1       ND          la      lb       2+3            ND
    73     GE5877                  L. weishimeri                     V/VI; VI: V1I;X: B;C; D          2            1        1          la      lb        ND            ND
    74     BE5963                  L. welshimeri                     V/VI ;VI: V1I;X: B; C ; D        2            1        1          la      lb        ND            ND
    75     BE6108                  L. weishimeri                     V/Vl;VI;VII;X; no H-factors      2            1        1          la      lb        ND            ND
    78     GE6254                  L. welshimeri                     V/VI ;VI: V1I;X: B; C ; D        2            1        1          la      lb        ND            ND
    84     LU6662                  L. welshimeri                     V/VI ;VI; VII ;X; B ;C; D        2            1       ND          la      lb       2+3            ND
    85     BE61 12                 L. welshimeri                     V/VI ;VI: VI1;X: B; C; D         2            1        1          la      lb        ND            ND
    86     AG7166                  L. welshirneri                    V/VI; VI: VI1;X: B; C; D         2            1        1          la      ib        ND            ND
    19     SLCC3954/ATCCC 359678   L. seeligeri                      1/2b                             2            2        1          lb      la        3              3
    20     8375                    L. seeligeri                      4ab                              2            2        1          lb      la        3              1
    21     BE8950                  L. seeligeri                      1/2b                             2            2        1          lb      ND        3              1
    22     BE8953                  L. seeligeri                      4d                               2            2        1          lb      ND        3              2
    64     NV2918                  L. seeligerid                     1/2b                             2        2       m          l     b      ND        3             ND
    27     LL278                   L . ivanovii subsp. ivanovii      5                                2            2        1          2a      la        6             ND
    28     LL483                   L. ivanovii subsp. ivanovii       5                                2            2        1          2a      la        6             21
    29     BE5087                  L . ivanovii subsp. ivanovii      5                                2            2        1          2a      ND        6             21
    46     CLIP 125lOT§            L. ivanovii subsp. ivanovii       5                                2            2        1          2a      ND        6             21
    51     SLCC4306                L . ivanovii subsp. ivanovii      5                                2            2        1          2a      ND        6             20
    55     SLCC2098                L. ivanovii subsp. ivanovii       5                                2            2        1          2a      ND        6             21
    30     BE1604                  L. ivanovii subsp. londoniensis   5                                2            2        1          2b      la        4             19
    31     BE5063                  L. ivanovii subsp. londoniensis   5                                2            2        1          2b      la        ND            ND
    32     BE1694                  L. ivanovii subsp. londoniensis   5                                2            2        1          2b      ND         4            ND
    33     BE3728                  L. ivanovii subsp. londoniensis   5                                2            2        1          2b      ND         4            ND
    47     CLIP12229T5             L. ivanovii subsp. londoniensis   5                                2            2        1          2b      la         4            18
    48     BE5194                  L. ivanovii subsp. londoniensis   I;I/II;A;B;C (1/2b)              2            2        1          2b      ND         4            ND
    50     BE5195                  L. ivanovii subsp. londoniensis   I/II;VII;no H-factors            2            2        1          2b      ND        ND            ND

                                                                                                                                                              continued opposite

130                                                                                                         International Journal of Systematic Bacteriology 48
                                                                                         PCR fingerprinting for identification of Listeria species

Table 1 (cont.)

I   52
                Original no.


                                L. ivanovii subsp. londoniensis
                                L. ivanovii subsp. londoniensis
                                                                         Serotype or unusual
                                                                       agglutination reactions1

                                                                   I; I/II;A; B ; C (1/2b)








    43   ATCC 2540211           L. grayi                                                           3      2    2       2a          2         5             17
    44   ATCC 2540311           L. gra-vi                                                          3      2    2       2a          2         5             15
    45   SLCC2080/ATCC 191208   L . grayi                                                          3      2    2       2a          ND        5             16
    34   LA3056                 Brochothrix thermosphacta                                          4      3    2        1          ND       ND            ND
    35   LA3674                 Bacillus cereus                                                    5      2    3       2b         ND        ND            ND
    39   CIP 103610T            Carnobacterium gallinarum                                          9      6    4       4      N    D        ND            ND
    42   LA3193                 Cellulomonas (Oerskovia) turbata                                   11     8    6        ~          N        D
                                                                                                                                            ND            ND
    36   LA2365                 Enterrococcus durans                                               6      4    4       4      N    D        ND            ND
    37   LA2390                 Enterococcus jaecalis                                              7      4    4       4      N    D        ND            ND
    38   LA3225                 Jonesia (Listeria) denitri$cans                                    8      5    5       5          ND        ND            ND
    40   CIP 102976T            Vagococcusfluvialis                                                10     7    4       3          ND        ND            ND

ND,  Not determined.
* Superscript   letters indicate the following properties : a, arabitol-negative ; b, L-rhamnose-negative ; c, methyl a-D-mannoside-
negative ; d, methyl a-D-mannoside-positive.
t Serovar numbering and serological reactions are according to Seeliger & Hohne (35). Serotypes with flagellar factors B; C ; D or
somatic factors VII ;VIII or I ;1/11;VII have not been described previously.
$ rDNA-SSCP analysis was at room temperature.
§Isolates of which the 16s rRNA gene was sequenced in this study; their EMBL accession numbers are X98526X98532.
11 These isolates were formerly ‘L. murruyi’.

values are given in Table 3 and a tree representation of                               londoniensis and L. grayi) had specific and homo-
this calculation is shown in Fig. 2. The sequences                                     geneous profiles. The eight L. welshimeri isolates
obtained in the present study differ from those ob-                                    presented two different profiles, and for three of these
tained previously (7,8). The number of differences (not                                isolates this profile was identical to the unique L.
counting undefined bases as differences and comparing                                  innocua profile (Table 1). Electrophoresis was also
with the best corresponding to previously published                                    performed under different conditions (four different
sequences for each species) was between 3 bp (with L.                                  buffer systems, electrophoresis at 6 “C or at room
seeligeri sequence X56148) and 8 bp (with L. grayi                                     temperature, addition of 0 or 10% glycerol). The
sequence X56150). To construct a tree based on the                                     profiles obtained varied markedly, depending on the
16s rRNA sequences of the Listeria species studied                                     conditions used, but some or all L. welshimeri isolates
here, they were imported into the ARB package and                                      remained indistinguishable from L. innocua, whereas
aligned manually to match the alignment of the                                         differentiation of other species became difficult or
Listeria sequences determined previously (7, 8). A                                     impossible (data not shown).
phylogenetic tree containing the published sequences
and a combined tree containing also the sequences                                      rDNA-spacer-PCR
obtained in this study were constructed (data not                                      Analysis of the PCR products by PAGE showed that
shown). The clustering of the species in all three trees                               21 profiles could be found among the 37 isolates tested
did not show significant differences.Thus, the sequence                                (Fig. 4, Table 1). Several profiles were found within
discrepancies found between published and new se-                                      each species, except for L. welshimeri (for which only
quences had no influence on the determination of                                       two isolates were tested). The profiles were composed
phylogenetic relationships, although there were less                                   of one main band and one or several minor bands. The
interspecific differences for the newly determined                                     variation within a species relied on both the minor and
sequences. Computer-aided restriction digestion of the                                 the main bands. Because of the variability within the
newly determined sequences corresponded perfectly                                      species it was difficult to define reliable diagnostic
with the observed ARDRA patterns, except for two                                       criteria for species identification. rDNA-spacer-PCR
minor discrepancies (Table 2).                                                         was clearly reproducible. When repeating electro-
                                                                                       phoresis with new amplification products from the
rDNA-SSCP analysis                                                                     same strains, the same pofiles were obtained and the
                                                                                       same differences between profiles were reproducibly
The results of the rDNA-SSCP analysis are reported in                                  detectable.
Table 1 and are illustrated in Fig. 3. Seven profiles were
obtained for the six Listeria species examined. The                                    t DNA-spacer-PCR
profiles were composed of two or four bands. Five
Listeria species and subspecies ( L . monocytogenes, L.                                The discriminatory power of tDNA-spacer-PCR de-
seeligeri, L. ivanovii subsp. ivanovii, L. ivanovii subsp.                             pended highly on the resolving power of the electro-

International Journal of Systematic Bacteriology 48                                                                                                             131
Table 2. Comparison of observed ARDRA patterns with restriction fragment sites and lengths as predicted by 165 rRNA sequence determination

  Restriction with                                                     Restriction sites                                                                                               Restriction fragment lengths                                                ARDR
  AIuI (AG/CT)
                            3213       127             140    207      822/3   98213 1029/30            1234/5                   13         3213        47       67      95      160   175    205     23213   438          61516

 L. monocytogenes            +                                 +        +        +           +       +                                       +          +                         +    +      +           +   +
 L. innocua                  +                                 +        +        +           +                                               +          +                         +     +                     +        +
 L. welshimeri               +                                 +        +        +           +                                               +          +                         +     +                     +        +
 L. seeligeri                +                                 +        +        +           +                                               +          +                         +     +                     +        +
 L. ivanovii ivanovii        +                                 +        +        +           +                                               +          +                         +     +                     +        +
 L. ivanovii londoniensis    +                                 +        +        +           +                                               +          +                        +      +                     +        +
 L. gmyi                     +             +           +       +        +        +           +       +                           +          +          +         +       +       +             +          +            +
 CfOI (CCCIC)               145 53718l9140 723/4             1069170                                                             2          145        182       346     392     395         53718

 L. monocytogenes            +         ++               +      +                                                                -       H          +         +       +       +    +                                                                                 1
 L. innocua                  +         ++              +       +                                                                +       +          +         +    +       +      +                                                                                  1
 L. welshimeri               +     -           H       +       +                                                                +       +          +         +    +       +      +                                                                                  1
 L. seeligeri                          -       H       +       +                                                                ++                      +        +               +      +                                                                           2
 L. ivanovii ivanovii                  ++              +       +                                                                ++                      +        +               +      +                                                                           2
 L. ivanovii Iondoniensis              ++              +       +                                                                ++                      +        +               +      +                                                                           2
 L. gmyi                            +          +   t    +      +                                                                ++                      +        +               +      +                                                                           2

 BsiUI (CGICG)              196        362             487    537      539     72314       93213   121718 1224/ 131112 13261     2          718         15               50      87    94    125      14011   166     18415    196/7   209   24213   285   29213
                                                                                                            5            7

 L. monocytogenes            +         +               +       +        +        +          +       +        +                   +           +                           +                    +               +        +       +       +      +      +              la
 L. innocua                  +         +               +       +        +        +          +       +       +                    +          +                            +                    +               +        +       +       +      +      +              la
 L. welshimeri               +         +               +       +        +        +          +               +                    +                                       +                    +               +        +       +       +      +             +       la*
 L. seeligeri                +         +               +       +        +        +          +               +                   +                                        +                    +               +        +       +       +      +             +       lb
 L. ivanovii ivanovii        +         +               +       +        +        +          +       +       +    +      +       +           +           +                +       +            +           +    +       +           +   +             +              2a
 L. ivanovii londoniensis   +          +               +       +        +        +          +               +    +      +       +                       +                +       +            +           +    +       +           +   +                    +       2b
 L. gmyi                    +          +               +       +        +        +          +       +            +      +       +                       +                +             +      +           +    +       +           +   +             +              2a

 nu91 (TITAA)               553    579/80              831    837      917      919    1051/2      1076                        25/6/7   8216           13314 252         257     391   415          553

 L. monoqvtogenes           +           +              +                +                   +                                    +          +          +         +                     +      +                                                                     la
 L. innocua                 +           +              +                +                   +                                    +          +          +         +                     +      +                                                                     la
 L. welshimeri              +           +              +                +                   +       t                           +       t         +          +       +           +            +                                                                     lb
 L. seeligeri               +           +              +                +                   +                                    +          +          +         +                     +      +                                                                     la
 L. ivanovii ivanovii       +           +              +                +                   +                                    +          +          +         +                     +      +                                                                     la
 L. ivanovii londoniensis   +          +               +                +                   +                                    +          +          +         +                     +      +                                                                     lb
 L. grayi                   +          +                       +                 +          +                                    +          +           +                +             +      +                                                                     2t

*According to the sequences determined in this study, L. welshimeri has a deletion which destroys the BstUI restriction site at position 1217. This predicts ARDRA pattern lb. However, for all L. welshimeri
strains ARDRA pattern 1a was observed.
?For L. gruyi, both the 252 and 257 bp fragments were observed on ARDRA patterns instead of the predicted 257 bp fragment only.
                                                                                                                                                                    PCR fingerprinting for identification of Listeria species

 Table 3. Distance values data matrix of the Listeria 165 rRNA gene sequences determined in this work
Values were calculated using the neighbour-joining method (Kimura correction, vertical gaps compressed) provided by the ARB
software package.

                                                                                        L. grayi L. innocua                                           L. ivanovii subsp.                              L. monocytogenes                                L. seeligeri L. welshimeri

                                                                                                                                                    ivanovii        londoniensis

   L. grayi                                                                               0.0000                      0.0424                        0.0348                0.0393                                   0-0386                                   0.0393                           0.0401
   L. innocua                                                                                                         0.0000                        0.0 100               0.0093                                   0.0057                                   0-0071                           0-0057
   L. ivanovii subsp. ivanovii                                                                                                                      0.0000                0.0035                                   0.0 122                                  0.0064                           0.0 100
   L. ivanovii subsp. londoniensis                                                                                                                                        0-0000                                   0.0 136                                  0.0043                           0.0085
   L. monocytogenes                                                                                                                                                                                                0~0000                                   0.0114                           0.0 100
   L. seeligeri                                                                                                                                                                                                                                             0~0000                           0.0057
   L. welshimeri                                                                                                                                                                                                                                                                             o*oooo

                                                   L. ivanovii                                                                                                                                                .....................................................................................................
 L.   ivanovii
 subsp. londoniensis subsp. ivanovii                                                                                                                                                                          Fig. 2. Tree representation of distances
                                                                                                                                                                                                              between the 165 rDNA sequences of Listeria
                 L. seeligeri                                                                                                                                                                                 species. Only the sequences as determined in
                                                                                                                                                                                                              this study were used for the above
                                                                                                                                                                                                              representation. E. coli K-12 was used as an
                  innocua                         L. welshimeri                                                                                                                                               outgroup. The tree was constructed with the
                                                                                                                                                                                                              built-in function of the ARB software
                L. monocytogenes                                                                                                                                                                              package using the          neighbour-joining
                                                                                                                                                                                                              method (Kimura correction, vertical gaps
                                                                                                                                                                                                              cpm-pressed). The value for the distance
                                                                                                                                                                                                              between E. coli and the Listeria cluster is
                                                                                        0.05                                                                                                                  indicated by the arrows.

                                                                                                                                                                 intraspecific differences were observed, several - inter-
                                                                                                                                                                 PCR and inter-electrophoresis run - reproducible dif-
                                                                                                                                                                 ferences between the species were present and enabled
                                                                                                                                                                 clustering into groups corresponding with established
                                                                                                                                                                 taxonomy (Table 4, Fig. 5). For instance, L. mono-
                                                                                                                                                                 cytogenes isolates were characterized by the presence
                                                                                                                                                                 of a 284, 285 or 286 bp peak, absent in all other
                                                                                                                                                                 isolates. Two L. monocytogenes isolates [SLCC 2374
                                                                                                                                                                 (LIS9) and SLCC 2376 (LIS1 l)] lacked this peak, but
                                                                                                                                                                 could be differentiated from all other species by the
                                                                                                                                                                 presence of a 265 and a 267 bp peak (Fig. 5). L. innocua
                                                                                                                                                                 showed most intraspecific heterogeneity and had most
                                                                                                                                                                 peaks in common with L. monocytogenes, from which
                                                                                                                                                                 it could only be differentiated by the absence of the
                                                                                                                                                                 284-286 and the 267 bp peaks. The profile of one L.
                                                                                                                                                                 innocua isolate (TI6520) showed combined charac-
                                                                                                                                                                 teristics of L. innocua, L. welshimeri and L. mono-
fig. 3. SSCP profiles of an amplified 216 bp fragment of the 165
rDNA as observed after an overnight run in non-denaturing
                                                                                                                                                                 cytogenes profiles (Table 4), and thus, despite repeated
0-5 x MDE gel in 0.5 x TBE +5 % glycerol at room temperature                                                                                                     testing, could not be identified. This is the same isolate
at 90 V and silver staining.                                                                                                                                     with a slightly aberrant BstUI restriction pattern
                                                                                                                                                                 (Table 1). All isolates of L. monocytogenes, L. innocua
                                                                                                                                                                 and L. welshimeri had a tRNA spacer of 159 bp, while
                                                                                                                                                                 this spacer was 160 bp long for all L. seeligeri isolates
phoresis technique used. Agarose gel electrophoresis                                                                                                             and 162 bp (with sometimes an additional 161 bp
and 5 % PAGE were not sufficiently discriminatory for                                                                                                            peak) for all L. ivanovii isolates. The subspecies of the
interpretation of the tDNA-spacer-PCR profiles (data                                                                                                             latter species could be differentiated by a single bp
not presented).                                                                                                                                                  length difference of a tRNA spacer region of 85 bp for
                                                                                                                                                                 L. ivanovii subsp. londoniensis and 86 bp for L. ivanovii
Only capillary electrophoresis provided sufficient res-                                                                                                          subsp. ivanovii.-Isolates of species other than Listeria
olution to yield useful results. Although substantial                                                                                                            had clearly different tDNA PCR profiles.

International Journal of Systematic Bacteriology 48                                                                                                                                                                                                                                                             133
M. Vaneechoutte and others

.....,.....,.....,.................,...........................,,...........................................,..............................................................................,........,..............,.,...........,.,............................,.........,... .................................

              Fig. 4. rDNA-spacer-PCR profiles as observed after 5 % PAGE and ethidium bromide staining (negative image).

For tDNA spacer amplification, reproducibility prob-                                                                                                           This is particularly true for L. monocytugenes and L.
lems of the amplification step were observed. Upon                                                                                                             innocua, which must be clearly distinguished because
analysis with agarose gel electrophoresis and PAGE,                                                                                                            of the health hazard represented by the former but not
sometimes additional peaks were observed depending                                                                                                             by the latter. However, both the haemolysis and the
on the PCR run and relative peak height was variable                                                                                                           Christie-Atkins-Munch-Petersen (CAMP) test with
between PCR runs (data not shown). Similar problems                                                                                                            Staphylococcus aureus, which are the distinctive cri-
have been observed with e.g. arbitrarily primed PCR                                                                                                            teria for these two species, are not always easy to
(36). However, highly identical profiles for different                                                                                                         interpret because some L. monocytogenes isolates
PCR runs were observed using capillary electro-                                                                                                                produce only very faint haemolysis and/or weakly
phoresis.                                                                                                                                                      positive CAMP reaction. A positive CAMP reaction
                                                                                                                                                               with Rhodococcus equi has recently been proposed as
Aty picaI isolates                                                                                                                                             an additional identification criterion for L. mono-
                                                                                                                                                               cytogenes (11). However, this test has long been
The molecular identification of atypical L. ivanovii                                                                                                           controversial and has not been adopted by all lab-
subsp. londoniensis and of all the atypical L. mono-                                                                                                           oratories. Its validity also heavily depends on the R.
cytogenes and L. seeligeri using ARDRA, rDNA-                                                                                                                  equi strain used for the reaction (1 1).
SSCP analysis (Table I), and tDNA-spacer-PCR
analysed on capillary electrophoresis was in full                                                                                                              Several commercial galleries have recently been de-
agreement with the results obtained with the API                                                                                                               veloped to make the identification of Listeria species
galleries (Table 1). The phenotypically atypical L.                                                                                                            easier (1, 2, 24). Nevertheless, since some identifi-
innocua and L. welshimeri isolates could only be                                                                                                               cations still rely on a unique characteristic (2) for
definitely identified by ARDRA and tDNA-spacer-                                                                                                                which some isolates may show an ambiguous reaction,
PCR.                                                                                                                                                           additional tests like haemolysis and CAMP may still
                                                                                                                                                               be necessary for a correct identification (24). Fur-
                                                                                                                                                               thermore, biochemically abnormal isolates are regu-
                                                                                                                                                               larly encountered during routine identification of
Because of their high phenotypic similarity, only a                                                                                                            Listeria isolates. Thus, in spite of tremendous improve-
limited number of biochemical tests can be used to                                                                                                             ments in commercial kits, phenotypic identification
differentiate the Listeria species from one another (3 1).                                                                                                     still remains difficult for a proportion of the isolates (1,

134                                                                                                                                                                                          International Jo urna I of Systematic Bacteriology 48
                                                                                                                         PCR fingerprinting for identification of Listeria species

                                                                                                                        2). In such cases, rapid and reliable confirmation based
                                                                                                                         on genetic criteria remains of great interest.
                                                                                                                        We evaluated the effectiveness of four of the PCR
                        + + > + I                                                       I           I       I       I
                                                                                                                        fingerprinting techniques for the identification of
                                                                                                                        isolates of the genus Listeria. ARDRA enabled differ-
                            I           I               I           I       I           I           I       IF          entiation among all six species and between the two L .
                                                                                                                        ivanovii subspecies, although only minor differences
                                                                                                                        were detected between the non-haemolytic species L .
                                                                                                                        innocua and L. welshimeri and between both L . ivanovii
                            I           I           S           1           I           I       I           I       I   subspecies.
                                                                                                                        High discriminatory power was also obtained with
                                                                                                                        rDNA-SSCP analysis (44), which identified all the
                                                                                                                        Listeria species, except L. innocua and some L.
                            I           I               I           I       I           I + + +                         welshirneri isolates, which shared indistinguishable
                                                                                                                        profiles under several of the electrophoresis conditions

                       1                                                                                                tested. Since the sequences of the 16s rRNA gene

                        +           I               I           I       I           I       I           I       I
                                                                                                                        regions amplified for SSCP analysis diverge at only
                                                                                                                        one position for these two species, this lack of
                                                                                                I           l       l
                                                                                                                        discrimination is not unexpected. A more recent study
                                                                                                                        (45) showed that SSCP analysis of the rDNA required
                            I       +           I
                                                                                                I           l       l
                                                                                                                        two different primer sets to differentiate all six Listeria
                                                                                                                        species. Since six copies of the rRNA-operon are
                            I           l           l                                           0
                                                                                                1                   I   present in the genome of Listeria species (27), it is
                                                                                                                        possible that the four bands observed in some of the L.
                        +       l           +               l               I           I   +           I           I   welshimeri isolates rely on the simultaneous presence
                                                                                                                        of two distinct alleles of the rRNA operon in the
                            I           I I + + + l                                                         I       I   genome of these particular isolates.



                       + + + +


                                                    l           I

                                                                            I I + + l

                                                                            I   +
                                                                                    l       I



                                                                                                                        Jensen et a/. (20) reported that Listeria isolates can be
                                                                                                                        identified on the basis of the sizes of the 16s-23s
                                                                                                                        intracistronic rDNA spacers. The molecular sizes of
                                                                                                                        the major fragments obtained by this method in the
                                                                                                                        present work were approximately the same as those
                                                                                                                        originally described. However, we found on several
                                                                                                                -       occasions differences in the banding profiles of isolates
                                                                                                                        belonging to the same species, sometimes even in the
                            I       I               I               I       I           I       I           I   +       major bands which have been suggested as diagnostic
                       *                                                                                                markers for species identification (Fig. 4). In our
                       + + + +                                              I +                 I           I       I   hands, this technique was therefore not useful for easy
                                                                                                                        identification of the species of the genus Listeria,




                                                                    I   +



                                                                                                        +       I
                                                                                                                        because of the lack of species-specific bands within the
                                                                                                                        profiles of some species and because sometimes only
                                                                                                                        minor electrophoretic migration differences were ob-
                       + + * I                                      I   +           I           I           I       I   served between profiles of different species. The reason
                            I       I               I               I       I       I           I           IF
                                                                                                                        for the discrepancy between both studies is not known,
                                                                                                                I       but the differences may be due to the use of different
                                                                                                                        populations of Listeria isolates. Although we ex-
                                                                                                                        amined less isolates than Jensen et al. (20), the isolates
                                                                                                                        used here were selected to cover a large spectrum of
                                                                                                                        serotypes and may represent a more heterogeneous
                                                                                                                        population than that used by these authors. Our results
                                                                                                                        suggest that the sequence and length variability of the
                                                                                                                        rDNA spacers may be greater than expected in the
                                                                                                                        genus Listeria and that, as in other species ( 5 , 14, 15,
                                                                                                                        25), rDNA-spacer-PCR may be used as a typing tool
                                                                                                                        rather than an identification tool. Jensen & Straus (21)
                                                                                                                        pointed to the need for using special amplification
                                                                                                                        conditions i.e. low concentration of polymerase and

                                                                                                                        polymerase buffer - to avoid multiple banding
                                                                                                                        patterns. However, we obtained identical profiles when
                        4 4 4 4 4 4 4 4 4                                                                               using these special conditions (data not shown).

International Journal of Systematic Bacteriology 48                                                                                                                           135
M. Vaneechoutte and others

                                                I                            I                                    Ill    1
                                                                                                                         I       LIS15inn
                                      I         I                            I                                      II   II      LIS17inn
                                                I                            1                                      II   II      LIS16inn
                                                1                            1                                      II   I1      LIS18inn
                                                I                            I                                      II    I      LIS25inn
                                                I                          I                                        II    I      LIS23wel
                                                I                          I                                        II   I
                                                I                         I                                      II II   I       LI52 1see
                                                I                         I                                      111     I       LIS22see
                                                I                         I                                       II     I       LIS19see
                                                 1                       I                                          II    I      LIS28iviv
                                                 I                       I                                         I1 I          LIS29iviv
                                                 I                       I                                         II II         LIS27iviv
                                                 I                       I                                         I1 I          LIS3Oivlo
                                                 I                       I                             i        II         I     LIS3livlo
                                      I              I                       I                    I           Ill        II
                                      I              I                       I                    I           Ill          I     LIS13mon
                                      I              I                       I                    I           Ill          II    LlS7mon
                                      I              I                       I                    I           Ill          I     LIS14mon
                                      I              I                       I                    I           Ill         II     LIS5mon
                                      I              I                       I                    I           Ill        I1      LIS6mon
                                      I              I                           I                1           Ill         II     LIS8mon
                                      I              I               I           I   I            I           IIII       II      LlSl2mon
                                      I              I                           I                I           Ill         Ill    LIS3mon
                                      I              I                           I                II           Ill         II    LIS4mon
                                      I              I                           I                             Ill        111    LlSlmon
                                      I              I                           I                             Ill         111   LlSZmon
                                      I         I                                I                           IIII         II     LIS9mon
                                      I         I                                I                             Ill       II      LISllmon

      Fig. 5. Clustering with Ward algorithm of tDNA-spacer-PCR profiles obtained by capillary electrophoresis on an ABI 310
      apparatus. Calculation of similarity matrix based on band-matching using area-sensitive coefficient (tolerance band
      position 0.1%, minimal surface bands: 0 % of total profile). Similarity indicated as a percentage on top of the
      dendrogram; fragment length of amplified tRNA spacers indicated in bp on top of fingerprints. Abbreviations of strain
      designations: inn, L. innocua; iviv, L. ivanovii subsp. ivanovii; ivlo, L. ivanovii subsp. londoniensis; mon, L.
      monocytogenes; see, L. seeligeri; wel, L. welshimeri. Numbers in the strain designation abbreviations refer to the study
      code as indicated in Table 1.

Capillary electrophoresis could reproducibly separate                hand, L. ivanovii and L. grayi have species-specific
fragments differing by only a single base pair in length             antigenic combinations and the biochemically very
and only this high resolution enabled sufficient dis-                close species L. rnonocytogenes and L. innocua do not
crimination between the tDNA PCR profiles. En-                       share the same serovars and can therefore be differ-
hanced tDNA-PCR reproducibility by the use of                        entiated serologically. On the other hand, L. seeligeri
capillary electrophoresis may be explained by the fact               cannot be serologically distinguished from L . mono-
that the high resolution offered by this electrophoresis             cytogenes, nor can L. welshimeri be distinguished from
technique enables the detailed analysis of smaller sized             L. innocua.
fragments (basically in the range 60-300 bp), which                  The species identification of 13 antigenically abnormal
may be more reproducibly amplified.                                  Listeria isolates (Table I), as obtained with standard
However, the subtlety of the observed differences, the               biochemical tests and with the API Listeria galleries,
peak height variability, the possible interference of                was confirmed by ARDRA, rDNA-SSCP analysis and
background amplification products and the huge                       tDNA spacer analysis. Among others, these data show
intraspecific variability (as was also observed with 5 %             that serovar 1/2b isolates can also be found in Listeria
PAGE) require careful evaluation of the practical                    species other than L. rnonocytogenes and L . seeligeri
applicability of tDNA-spacer-PCR. It can be con-                     (i.e. in L. innocua, L. ivanovii subsp. londoniensis and L.
cluded that thorough standardization of PCR con-                     welshirneri) and consequently, that serotyping may not
ditions, high-resolution electrophoresis, and powerful               be as trustworthy, as is generally accepted, as a
pattern recognition and clustering software are needed               confirmatory tool for species identification. In ad-
to automate species identification based on tDNA-                    dition, PCR-based DNA fingerprinting enabled the
spacer-PCR.                                                          identification of four rarely encountered biochemical
In addition to biochemical characterization, Listeria                reaction patterns in L . rnonocytogenes and L. seeligeri
isolates may also be differentiated on the basis of their            (indicated as a-d in Table 1).
antigenic composition. Thirteen somatic antigens and                 Because of discrepancies between 16s ARDRA results
five flagellar antigens have been described in the                   and computer-assisted restriction analysis of the pub-
Listeria species and are currently used for serotyping               lished sequences (7,8), the sequences of the 1 s rRNA
(35). However, undescribed combinations of antigens                  genes of seven representative isolates were determined
are regularly encountered during serotyping of field                 again. Computer-assisted restriction analysis of the
isolates, several of which are listed in the present work.           sequences as determined in this study was almost in
Some but not all serovars are species-specific. On one               complete agreement with ARDRA results (Table 2),

136                                                                                      International Journal of Systematic Bacteriology 48
                                                            PCR fingerprinting for identification of Listeria species

thus confirming the validity of the new updated rRNA       the reliable identification of biochemically and sero-
gene sequences.                                            logically atypical Listeria isolates. The data reported
Besides possible sequencing errors, other reasons have     here confirm a high genetic relatedness between most
been discussed previously (6) and may explain the          Listeria species and suggest a closer relationship than
discrepancies between the three independant 16s            expected between the two non-haemolytic species L.
rRNA gene sequence determinations carried out thus         innocua and L. welshimeri, which also share common
far (7, 8, this study). Furthermore, application of        serotypes. The combination of the 16s rDNA sequence
reverse transcription prior to sequencing may be a         data and the data gathered with the different PCR
major source of sequence errors.                           fingerprinting methods suggests that L. innocua is as
                                                           closely related to L. welshimeri as it is to L. mono-
The results obtained here with different PCR-based         cytogenes and confirms that there are subtle but
DNA fingerprinting techniques reflect the previously       consistent genetic differences between the two L.
established high intragenic relatedness of the Listeria    ivanovii subspecies and that L. grayi is homogeneous
species. For example, for a total of 17 restriction        and is validly included in the genus Listeria.
enzymes out of 22 tested, not a single restriction site
difference could be established in the 16s rDNA of the
six species. Correspondingly, the new sequences pre-       ACKNOWLEDGEMENTS
sented here indicate that the dissimilarity estimates      We are very grateful to C . L. Gyles for checking the
between the 16s rRNA genes of the different Listeria       manuscript.
species are even lower than previously reported (7),
although this does not influence the phylogenetic
relatedness as has been established previously.            REFERENCES
Subtle, but reproducible, differences between L.            1. Bannerman, E., Yersin, M. N. & Bille, 1. (1992). Evaluation of
ivanovii subsp. ivanovii and L. ivanovii subsp. london-        the Organon-Technika MICRO-ID LISTERIA system.
iensis could be established by several techniques.             Appl Environ Microbiol58, 201 1-2015.
Sequence comparison of the first 1200 bp of the 16s         2. Bille, J., Catimel, B., Bannerman, E., Jacquet, C., Yersin, M. N.,
rDNA revealed only two base pair substitutions (at             Caniaux, I., Monget, D. & Rocourt, J. (1992). API Listeriu, a
positions 46 and 1096) between both subspecies, and            new and promising one-day system to identify Listeriu
                                                               isolates. Appl Environ Microbiol58, 1857-1 860.
single bp length differences were observed between
tDNA spacers of isolates of both subspecies. These          3. Boerlin, P., Rocourt, J., Grimont, F., Grimont, P. A. D., Jacquet,
                                                               C. & Piffaretti, I.-C. (1992). Listeriu ivunovii subsp. london-
data fit the classification of the two corresponding
                                                               iensis subsp. nov. Int J Syst Bucteriol42, 69-73.
genomic groups at the subspecies level (3) and not as
different species. Only ARDRA and tRNA-spacer-              4. Boerlin, P., Rocourt, 1. & Piffaretti, I.-C. (1991). Taxonomy of
PCR could establish some differences between L.                the genus Listeriu by using multilocus enzyme electro-
                                                               phoresis. Int J Syst Bucteriol41, 59-64.
innocua and L. welshimeri. Interestingly, L. innocua
and L. welshimeri also share common serotypes. 16s          5. Cartwright, C. P., Stock, F., Beekmann, 5. E., Williams, E. C. &
                                                               Gill, V. J. (1995). PCR amplification of rRNA intergenic
rDNA sequence similarity distances as calculated here
                                                               spacer regions as a method for epidemiological typing of
for the species pairs L. monocytogenes - L. innocua            Clostridium dificile. J Clin Microbiol33, 184-1 87.
and L. welshimeri L. innocua are comparable (Table

3). Finally, the fingerprints obtained for L. grayi         6. Clayton, R. A., Sutton, G., Hinkle, P. S., Jr, Bult, C. & Fields, C.
                                                               (1995). Intraspecific variation in small-subunit rRNA se-
differed most from those of all the other Listeria             quences in GenBank. Why single sequences may not
species, thus confirming its distance from the rest of         adequately represent prokaryotic taxa. Int J Syst Bucteriol
the genus (4, 7, 12, 31). L. grayi was found to be             45,595-599.
homogeneous with all techniques (except with rDNA-          7. Collins, M. D., Wallbanks, S., Lane, D. J., Shah, J., Nietupski,
spacer-PCR), which is in agreement with the recent             R., Smida, J., Dorsch, M. & Stackebrandt, E. (1991). Phylo-
grouping of L. grayi and ‘ L .murrayi’ into one unique         genetic analysis of the genus Listeriu based on reverse
species (33).                                                  transcriptase sequencing of 16s rRNA. Int J Syst Bucteriol
                                                               41, 24&246.
Conclusions                                                 8. Czajka, J., Bsat, N., Piani, M., Russ, W., Sultana, K., Wiedmann,
                                                               M., Whitaker, R. & Batt, C. A. (1993). Differentiation of
In conclusion, the previously published 16s rRNA               Listeriu monocytogenes and Listeriu innocuu by 16s rRNA
gene sequences of all six Listeria species have been           genes and intraspecies discrimination of Listeriu mono-
updated and now fit the restriction profiles obtained by       cytogenes strains by random amplified polymorphic DNA
ARDRA. In our hands, ARDRA performed best with                 polymorphisms. Appl Environ Microbiol59, 304308.
respect to both discriminatory power and practical          9. Deng, S., Hiruki, C., Robertson, 1. A. & Stemke, G. W. (1992).
applicability. Satisfactory results were obtained with         Detection by PCR and differentiation by restriction frag-
rDNA-SSCP analysis and tDNA-spacer-PCR, al-                    ment length polymorphism of Acholeplusmu, Spiroplusmu,
though the former requires thorough standardization            Mycoplusmu, and Ureuplusmu, based upon 16s rRNA
of electrophoresis conditions and the use of additional        genes. PCR Methods Appl 1,202-204.
primer sets, and the latter requires high-resolution       10. Ehrenstein, B., Bernards, A. T., Dijkshoorn, L., Gerner-Smidt,
electrophoresis. All of these techniques also allowed          P., Towner, K. J., Bouvet, P. 1. M., Daschner, F. D. & Grund-

International Journal of Systematic Bacteriology 48                                                                           137
M. Vaneechoutte and others

    mann, H. (1996). Acinetobacter species identification by             25. Kostman, J. R., Edlind, T. D., Lipuma, J. 1. & Stull,T. L. (1992).
    using tRNA spacer fingerprinting. J Clin Microbiol 34,                   Molecular epidemiology of Pseudomonas cepacia deter-
    24 14-2420.                                                              mined by polymerase chain reaction ribotyping. J Clin
11. Fernandez-Garayzabal, J. F., Suarez, G., Blanco, M. M.,                  Microbiol30, 20842087.
    Gibello, A. & Dominguez, L. (1996). Taxonomic note: a                26. McClelland, M., Petersen, C. & Welsh, 1. (1992). Length
    proposal for reviewing the interpretation of the CAMP                    polymorphisms in tRNA intergenic spacers detected by
    reaction between Listeria monocytogenes and Rhodococcus                  using the polymerase chain reaction can distinguish strepto-
    equi. Int J Syst Bacteriol46, 832-834.                                   coccal strains and species. J Clin Microbiol30, 1499-1504.
12. Feresu, 5. B. & Jones, D. (1988). Taxonomic studies on               27. Michel, E. & Cossart, P. (1992). Physical map of the Listeria
    Brochothrix, Erysipelothrix, Listeria, and atypical lacto-               monocytogenes chromosome. J Bacterioll74, 7098-7 103.
    bacilli. J Gen Microbiol 134, 1165-1 183.                            28. Nesme, X., Vaneechoutte, M., Orso, S., Hoste, B. & Swings, J.
13. Gilrtler, V., Wilson, V. A. & Mayall, B. C. (1991). Classification       (1995). Diversity and genetic relatedness within genera
    of medically important clostridia using restriction endo-                Xanthomonas and Stenotrophomonas using restriction endo-
    nuclease site differences of PCR-amplified 16s rDNA. J                   nuclease site differences of PCR-amplified 16s rDNA. Syst
    Gen Microbioll37, 2673-2679.                                             Appl Microbiol18, 127-135.
14. Gurtler, V. (1993). Typing of Clostridium dzjicile strains by        29. Ralph, D., McClelland, M., Welsh, J., Baranton, G. & Perolat,
    PCR-amplification of variable length 16s-23s rDNA spacer                 P. (1993). Leptospira species categorized by arbitrarily
    regions. J Gen Microbioll39, 3089-3097.                                  primed polymerase chain reaction (PCR) and by mapped
15. GUrtler, V. & Stanisch, V. A. (1996). New approaches to                  restriction polymorphism in PCR-amplified rRNA genes. J
    typing and identification of bacteria using the 16s-23s                  Bacterioll75, 973-98 1.
    rDNA spacer region. Microbiology 142, 3-16.                          30. Rocourt, J., Grimont, F., Grimont, P. A. D. & Seeliger, H. P. R.
16. Heyndrickx, M., Vandemeulebroecke, K., Scheldeman, P.,                   (1982). DNA relatedness among serovars of Listeria mono-
    Hoste, B., Kersters, K., De Vos, P., Logan, N. A., Aziz, A. M.,          cytogenes sensu lato. Curr Microbiol7, 383-388.
    Ali, N. & Berkeley, R. C. W. (1995). Paenibacillus (formerly         31. Rocourt, J., Schrettenbrunner, A. & Seeliger, H. P. (1983).
    Bacillus) gordonae (Pichinoty et al. 1986) Ash et al. 1994 is            Diffirenciation biochimique des groupes ginomiques de
    a later subjective synonym for Paenibacillus (formerly                   Listeria monocytogenes (sensus lato). Ann Microbiol Inst
    Bacillus) validus (Nakamura 1984) Ash et al. 1994: emended               Pasteur 134A, 65-7 1.
    description of P. validus. Int J Syst Bacteriol45, 661-669.          32. Rocourt, J., Wehmeyer, U., Cossart, P. & Stackebrandt, E.
17. Hookey, 1. V., Birtles, R. 1. & Saunders, N. A. (1995). Inter-           (1987). Proposal to retain Listeria murrayi and Listeria grayi
    genic 16s rRNA gene (rDNA)-23S rDNA sequence length                      in the genus Listeria. Int J Syst Bacteriol37, 298-300.
    polymorphisms in members of the family Legionellaceae. J             33. Rocourt, J., Boerlin, P., Grimont, F., Jacquet, C. & Piffaretti, 1.-
    Clin Microbiol33, 2377-238 1.                                            C. (1992). Assignment of Listeria grayi and Listeria murrayi
18. Jannes, G., Vaneechoutte, M., Lannoo, M., Gillis, M., Van                to a single species, Listeria grayi, with a revised description
    Canneyt, M., Vandamme, P., Verschraegen, G., Van                         of Listeria grayi. Int J Syst Bacteriol42, 171-174.
    Heuverswyn, H. & Rossau, R. (1993). Polyphasic taxonomy              34. Sallen, B., Rajoharison, A., Desvarenne, S., Quinn, F. &
    leading to the proposal of Moraxella canis sp. nov. for                  Mabilat, C. (1996). Comparative analysis of 16s and 23
    Moraxella catarrhalis-like strains. Int J Syst Bacteriol 43,             rRNA sequences of Listeria species. Int J Syst Bacteriol46,
    438-449.                                                                 669-674.
19. Jayarao, B. M., Dore, 1. J. E., Jr, Baumbach, G. A., Matthews,       35. Seeliger, H. P. R. & Htihne, K. (1979). Serotyping of Listeria
    K. R. & Oliver, 5. P. (1992). Restriction fragment length                monocytogenes and related species. Methods Microbiol 13,
    polymorphism analysis of 16s ribosomal DNA of Strep-                     3 1-49.
    tococcus and Enterococcus species of bovine origin. J Clin
    Microbiol30, 2235-2240.                                              36. van Belkum A., Kluytmans, J., van Leeuwen, W. &13 other
                                                                             authors (1995). Multi-center evaluation of arbitrarily
20. Jensen, M. A., Webster, 1. A. & Straus, N. (1993). Rapid                 primed PCR for typing of Staphylococcus aureus strains. J
    identification of bacteria on the basis of polymerase chain               Clin Microbiol33, 1537-1 547.
    reaction-amplified ribosomal DNA spacer polymorphisms.
    Appl Environ Microbiol59, 945-952.                                   37. Vaneechoutte, M., Cartwright, C. P., Williams, E. C., Jlger, B.,
                                                                             Tichy, HA., De Baere, T., De Rouck, A. & Verschraegen, G.
21. Jensen, M. A. & Straus, N. (1993). Effect of PCR conditions              (1996). Evaluation of 16s rRNA gene restriction analysis
    on the formation of heteroduplex and single-stranded DNA                 for the identification of cultured organisms of clinically
    products in the amplification of bacterial ribosomal DNA                 important Clostridium species. Anaerobe 2, 249-256.
    spacer regions. PCR Methods Appl3, 186-194.
                                                                         38. Vaneechoutte, M.,       De Beenhouwer, H., Claeys, C.,
22. Klmpfer, P., Bt)ttcher, S , Dott, W. & RUden, H. (1991).
                                .                                            Verschraegen, G., De Rouck, A., Paepe, N, Elaichouni, A. &
    Physiological characterization and identification of Listeria            Portaels, F. (1993). Identification of Mycobacterium species
    species. Int J Med Microbiol275, 423435.                                 with amplified rDNA restriction analysis. J Clin Microbiol
23. Kathariou, S. & Pine, L. (1991). The type strain(s) of Listeria          31,2061-2065.
    monocytogenes : a source of continuing difficulties.Int J Syst       39. Vaneechoutte, M., Dijkshoorn, L., Tjernberg, I., Elaichouni,
    Bacteriol41, 328-330.                                                    A., De Vos,       .
                                                                                              P, Claeys, G. & Verschraegen,          G. (1995).
24. Kerr, K. G, Rotowa, N. A., Hankey, P. M. & Lacey, R. W.
               .                                                             Identification of Acinetobacter genomic species by amplified
    (1991). Evaluation of the ROSCO system for the iden-                     ribosomal DNA restriction analysis. J Clin Microbiol 33,
    tification of Listeria species. J Med Microbiol35, 193-196.              11-15.

138                                                                                     International Journal of Systematic Bacteriology 48
                                                                         PCR fingerprinting for identification of Listeria species

40. Vaneechoutte, M., Riegel, P., de Briel, D., Monteil, H.,               polymorphisms in tRNA intergenic spacers for categorizing
   Verschraegen, G., De Rouck, A. & Claeys, G. (1995).                     staphylococci. Mol Microbiol6, 1673-1680.
    Evaluation of the applicability of amplified rDNA-restric-          44. Widjojoadmotjo, M. N., Fluit, A. C. & Verhoef, 1. (1994).
    tion analysis to identification of species of the genus                 Rapid identification of bacteria by PCR-single-strand
    Corynebacterium. Res Microbioll46, 633-641.                             conformation polymorphism. J Clin Microbiol 32,
41. Vaneechoutte, M., Rossau, R, DeVos, P., Gillis, M., Janssens,
                               .                                            3002-3007.
    D., Paepe, N., De Rouck, A., Fiers, T., Claeys, G. & Kersters, K.   45. Widjojoadmotjo, M. N., Fluit, A. C. & Verhoef, J. (1995).
    (1992). Rapid identification of bacteria of the Com-                    Molecular identification of bacteria by fluorescence-based
    amonadaceae with amplified ribosomal DNA-restriction                    PCR-single-strand conformation polymorphism analysis of
    analysis (ARDRA). FEMS Microbiol Lett 93, 227-234.                      the 16s rRNA gene. J Clin Microbiol33, 2601-2606.
42. Welsh, J. & McClelland, M. (1991). Genomic fingerprints             46. Wiedmann-Al-Ahmad, M., Tichy, H A . & Schan, G. (1994).
    produced by PCR with consensus tRNA gene primers.                       Characterization of Acinetobactev type strains and isolates
    Nucleic Acids Res 19, 861-866.                                          obtained from wastewater treatment plants by PCR finger-
43. Welsh, J. & McClelland, M. (1992). PCR-amplified length                 printing. Appl Environ Microbiol60, 4066-4071.

International Journal of Systematic Bacteriology 48                                                                               139